the salary weighted lottery today is a good step in the right direction IMO.

I think it has to be by industry and role. Software, mechanical engineering, teaching, medicine, etc are all very different

It needs to compensate for country and position.

It's the perfect model for open-source work because it's gonna end up in the training data anyway

I think it's overbloating my computer

It's a habit in specific countries, China is one in my experience.

It’s funny to me how I judge some cultural aspects as being worse. Of course it’s my own cultural bias, but I think the following things are “bad” for cultures to do:

1) long voice memos over succinctly typed notes

2) not forming lines/queues and just mobbing the train/shop/etc

3) habitually late or not on time

> 2) not forming lines/queues and just mobbing the train/shop/etc

Also even queueing can have different norms. Here people rush to a newly opened supermarket checkout like their life depended on it. I feel that should be a lot less sociably acceptable than it is, especially for those with overfilled carts that will add huge delays for anyone behind them.

Cursor was 60B. The 10B number was the breakup fee if the deal fell through.

$60B in SPCX stock, which is sort of magic money.

16 tokens / s is not nothing.

It's the other way around due to poor framing, it would be much easier to compare if you [the repo] said 0.02 tps.

Confusing numbers everywhere.

16tk/s... Then 3 tks per minute. Then someone else posted 0.3tk/s.

? Readme says 60-70s per token

Apple isn't counting on their model to be a frontier coding and cowork model. Gemini is perfectly fine for the tasks that new Siri is supposed to be doing.

That is not the Grok CLI being discussed. That's an open source, third party CLI. https://x.ai/cli is the official Grok CLI being discussed, and it is not open source.

thanks for the correction

The avg coding session has hundreds or thousands of tool calls. Even a 5% failure rate noticeably notches up token use and cost. See Gemini.

Yes, but each tool call has a different failure %. The tool calls that make up the majority of volume like grep are going to have nowhere near a 5% failure. A custom user-defined skill having a 5% failure rate is probably fine.

Doubt. Tried it, Grok 4.5 is leaps and bounds ahead of Gemini for real work.

And token costs vs. raw compute+electric cost is unit profitable too.